Research And Design Of Noninvasive Blood Viscosity Detection Methods And Systems

With the acceleration of life rhythm,unhealthy working hours and unhealthy eating habits have seriously affected people's health,and the incidence of cardiovascular disease and the number of patients show a clear upward trend.As one of the most important physiological parameters of the human body,blood viscosity is often used in the pathological diagnosis of cardiovascular diseases in clinic,and has important significance for the prevention,diagnosis and treatment of cardiovascular diseases.This article is based on the photoelectric volume pulse wave to carry out the non-invasive detection study of blood viscosity,in order to meet the health detection needs of the majority of cardiovascular disease population.Firstly,a hardware acquisition circuit and a filter circuit are designed.On this basis,signal preprocessing is performed on the photoelectric volume pulse wave.A FIR Butterworth filter and a wavelet-based denoising algorithm are designed to remove the high frequency noise and low frequency of the photoelectric volume pulse wave.Then the method of extracting the characteristic parameters of the photoelectric volume pulse wave is studied,which mainly includes the selection of the characteristic parameters and the research of the algorithm for extracting the characteristic parameters.This paper proposes a method for extracting the main peak and trough of the photoelectric volume pulse wave based on the Adam fast batch gradient ascent method.In addition,the wavelet mode minimum method is used to extract the complex repetitive wave peak and trough.The above extraction method is analyzed with the traditional differential method Comparison,the results show that the use of Adam fast batch gradient ascent method and wavelet mode minimum method is significantly better than the traditional differential method,which can effectively improve the extraction accuracy.Then,according to the photoelectric volume pulse wave characteristic parameters extracted above,human physiological parameters such as pulse wave waveform characteristic amount,vascular compliance,peripheral resistance,etc.are calculated to form a matrix.Using multivariate linear method and partial least squares method to fit blood viscosity and human physiological parameters,the blood viscosity prediction model is obtained,and an improved RBF neural network algorithm is proposed to construct the human physiological parameter-blood viscosity prediction model.By comparing the real blood viscosity with the prediction results of the prediction model,it is found by comparison that the prediction results of the improved RBF neural network model proposed in this paper are significantly better than the other two models.Finally,the non-invasive blood viscosity detection system is verified through experiments,and the test results are displayed on the host computer or cloud platform in real time,which verifies the feasibility of the non-invasive blood viscosity detection system.